Dynamic user profiles for web personalisation
Introduction
The dramatic growth of information on the WWW has inadvertently led to information overload and hence finding a specific piece of information has become difficult and time consuming (Challam, Gauch, & Chandramouli, 2007). Web personalisation systems have emerged in recent years in order to deal with this problem aiming to provide a personalised experience to users based on their individual preferences, interests and needs. Such systems have been developed for different domains of application (Pignotti et al., 2004, Challam et al., 2007, Sieg et al., 2007, Pan et al., 2007). In e-commerce, such systems have been used to recommend new items and products to users based on their previous purchasing history (Gorgoglione et al., 2006, Huang et al., 2004), while in e-learning, they are used to provide personalised e-learning services (Sun and Xie, 2009, Zhuhadar and Nasraoui, 2008). Personalisation systems require and maintain information about users, and this may include demographic data, interests, preferences, and previous history. One of the main challenges in such systems is that user interests, preferences and needs are not fixed, but change over time. If user profiles contain just static information, this eventually leads to constraining the personalisation process and recommending irrelevant services and items over time. To overcome this problem, methods are required for learning and understanding different user behaviours, and then adapting the profiles accordingly.
In this paper, we utilise ontological profiles to capture user interests and provide recommendations based on these. The contribution of our work is threefold. Firstly, we introduce two algorithms in order to improve the mapping process between web pages visited by the user that contain implicit information about the user interests and a reference ontology to explicitly represent these interests. Secondly, we introduce novel techniques to construct ontological short-term and long-term profiles that are tailored to the users, and adapt them based on their ongoing behaviour. Thirdly, the methods introduced attempt to recognise and handle potential interest drift and interest shift in the user interests.
To demonstrate our work, we introduce a personalisation system that consists of three phases. The first phase is the information retrieval phase which involves preparing a reference ontology, collecting user navigation behaviour, and mapping visited web pages to the reference ontology. Indeed, this phase is very important as capturing inaccurate user interests would directly affect the subsequent phases and eventually the personalisation performance. In this phase, we utilise two novel algorithms based on our work in Hawalah and Fasli (2011a) to improve the mapping process. The second phase is the profile adaptation and learning phase which utilises previous work in Hawalah and Fasli (2011b). This phase plays a major role in our model as it is responsible for learning, adapting and modelling ontological-user profiles. This also includes methods to adapt the ontological profiles to any shift or drift that might occur in the users’ behaviour. This phase makes use of a multi-agent system that coordinates the various processes and ensures that the user profile remains up-to-date. In the last phase, a re-ranking search system is introduced that utilises the dynamic user profile to provide a personalised search experience. The re-ranking search system takes advantage of the user interests to provide more personalised search results. To evaluate this work, we have conducted experiments with users over a period of time to assess the ability and effectiveness of our methods in tracking and adapting to changes in the user behaviour.
The rest of the paper is structured as follows. First we discuss related work. Section 3 presents the main architecture for modelling dynamic user profiles which consists of three phases with each phase being discussed in more detail in a subsequent section. We introduce the evaluation in Section 4. Section 5 describes the evaluation of the mapping and profile construction methods, while Section 6, details the evaluation of the dynamic user profiling methods in the context of their deployment within a personalised search system. The paper ends with the conclusions and pointers to future work.
Section snippets
Related work
As information on the WWW continues to proliferate, users find it increasingly difficult and time-consuming to sift through this information. To aid the user in his/her quest for the right information (be it on items, products, movies or articles), recommender and web personalisation systems have emerged. Recommender systems use two broad categories of techniques: content-based and collaborative-filtering (CF) techniques. The first technique views users as individuals. Such systems track the
Capturing and modelling dynamic user profiles
In this section, and following on from the identification of a number of drawbacks in existing works in personalisation, we propose a dynamic user profiling approach. In developing our research, we have taken into account a number of factors and desirable properties for personalisation systems:
- •
Users are reluctant to provide information about their interests and preferences explicitly through completing questionnaires, etc. (Montaner, 2001). Hence, a flexible personalisation system should try
Evaluation
In general, the evaluation of recommender, web personalisation or systems that deploy user profiles is recognised to be difficult and expensive (Yang, Padmanabhan, Rajagopalan, & Deshmukh, 2005). As such systems may have different purposes and they tend to be very complex, making direct comparisons is difficult. The evaluation strategies commonly used can be divided into three main categories (Shani & Gunawardana, 2011): offline, user-centred studies and online evaluations.
In offline
Evaluation I: mapping and user profile modelling methods
In setting up our user-centred study, we chose the domain of computers as the application domain. Our methods require the use of a reference ontology. Though we do not make any assumptions on the ontology used, for the purposes of this study, we first created a reference ontology using information extracted from the computer category from the Open Directory Project (ODP) Ontology. We used the ODP as it is considered to be the largest manually constructed directory consisting of websites
Evaluation II: dynamic user profiling for personalised search
We wish to evaluate the performance of the learning and adaptation processes in the context of a personalisation system. To this end, we implemented the proposed search personalisation system in Section 3.3 that utilises the dynamic user profiles to provide re-ranked search results based on user interests.
We used the same reference ontology that we created for the first evaluation and used Google7 as the search engine to retrieve the search results. As before, the Firefox
Conclusions
In this paper, we presented a dynamic user profile modelling approach for web personalisation. The starting point of our work was the identification of a number of issues in existing approaches, namely, the limitations of such systems in dealing with the user’s changing interests over time, the lack of a clear delineation of the short and long-term aspects of user interests and weaknesses in modelling dynamic aspects of the user’s behaviour. In our work, we aimed to address these issues.
Acknowledgements
Dr. Ahmad Hawalah was sponsored for his PhD studies by the University of Taibah, Medina, Kingdom of Saudi Arabia.
References (49)
- et al.
Exploring synergies between content-based filtering and spreading activation techniques in knowledge-based recommender systems
Information Sciences
(2011) - et al.
A semantic-expansion approach to personalized knowledge recommendation
Decision Support Systems
(2008) - et al.
Using ontology network analysis for research document recommendation
Expert Systems with Applications
(2008) - et al.
Generating semantically enriched user profiles for web personalization
ACM Transactions on Internet Technology
(2007) - et al.
Modern information retrieval: The concepts and technology behind search
(2011) - Basiri, J., Shakery, A., Moshiri, B., & Hayat, M. (2010). Alleviating the cold-start problem of recommender systems...
The IIR evaluation model: A framework for evaluation of interactive information retrieval systems
Information Research
(2003)- et al.
A multi-purpose ontology-based approach for personalised content filtering and retrieval
- Challam, V., Gauch, S., & Chandramouli, A. (2007). Contextual search using ontology-based user profiles. In Large scale...
- et al.
PVA: A self-adaptive personal view agent
Journal of Intelligent Information Systems
(2002)
Implicit interest indicators
Learning user interests for a session-based personalized search
A large-scale evaluation and analysis of personalized search strategies
Introducing semantics in web personalization: The role of ontologies
Ontology-based user profiling
User profiles for personalized information access the adaptive web
Interest drifts in user profiling: A relevance-based approach and analysis of scenarios
The Computer Journal
A graph model for E-Commerce recommender systems
Journal of the American Society for Information Science and Technology
Ontology-based user preference modeling for enhancing interoperability in personalized services
Cited by (58)
Dynamic user profile construction and its application to smart product-service system design: A maternity-oriented case study
2024, Expert Systems with ApplicationsKnowledge-driven profile dynamics
2024, Artificial IntelligencePersonalized information recommendation model based on context contribution and item correlation
2019, Measurement: Journal of the International Measurement ConfederationCitation Excerpt :However, in personalized services, the influence of user’s internal and external contexts on recommendation process should be taken into consideration, and the traditional recommendation model has declined in accuracy due to a lack of contexts [1,8,9]. Therefore, most conventional recommendation algorithms have tried to introduce contexts into recommendation process, but their user interest models have low accuracy [10,11]. They usually use simple models or even the user registration information for the interest extraction.
Time-Dependent Enhanced Personalized Semantic Map-Based Transportation Mode Recommendation System
2024, Lecture Notes in Networks and SystemsHYBRID PERSONALIZED RECOMMENDATION MODEL FOCUS ON IMPROVED COLLABORATIVE FILTERING
2023, Journal of Theoretical and Applied Information Technology